Seismic reflection surveys can reveal many structural details about a subterranean formation, including the location of subterranean faults, mineral deposits, and the like. It is desirable to accurately image and model the location and extent of these geologic structures owing to their importance in a number of commercial applications. For example, in hydrocarbon exploration, it is important to accurately model salt bodies and similar structures because such salt bodies are known to trap significant amounts of oil and gas in the formation underneath.
A common and widely used method of generating 3-D images of a salt body from seismic reflection surveys is to define the salt boundaries using horizons or height fields. Typically, an upward-facing or top horizon and a downward-facing or bottom horizon are defined for the salt body, then the salt structure between the top and bottom horizons is filled in by performing a flood fill. The data representing the horizons is usually stored and processed by imaging software using a 2-D array or grid where the elements in the grid represent points on the surface of the salt body in the horizontal direction (i.e., X and Y axes), and the value contained in each element indicates the depth (i.e., Z axis) of the salt boundary at that point.
However, constructing an accurate and realistic model of a salt body is inherently difficult because the nature of salt makes the seismic data noisy and poorly defined. In most cases, geologists and geophysicists must interpret the volumes of seismic data using their geological knowledge and experience to manually define the edge of the salt body as intersected by an individual vertical plane (section) and horizontal plane (slice). These seismic interpretations typically contain data points that were deemed by the geologists and geophysicists as most indicative of the boundary of the salt body. The data points are then input into imaging software, which connects the points together to form a set of polylines that outline the contour of the salt body. The imaging software then fills in the area between the polylines using the 2-D array or grid to render a 3-D image of the salt body.
Because salt bodies are closed structures, the polylines almost always encircle the salt body and are therefore almost always closed-ended. This means virtually every element in the 2-D array or grid for the image of a salt body will have at least two values in Z, with some elements having as many as four or more Z values, depending on the shape of the salt body. In addition, using polylines to describe the salt body can lead to inconsistencies (“mis-ties”) at the point where two polylines are intended to intersect. Additionally, editing any one polyline can introduce even more mis-ties to other polylines. These mis-ties typically cause the rendering of the salt body to breakdown. Furthermore, in order to correct a mis-tie, editing has to be performed on both polylines, which can be very time consuming.
A need therefore exists for improved techniques to overcome the deficiencies in using polylines to describe the salt bodies by automatically eliminating the mis-tie that may occurs at the intersection of the polylines.